Cracking the code: real-time monitoring of wood drying and the occurrence of cracks

Abstract

Predicting kiln-drying schedules for a given wood species is not yet possible, despite many research efforts. This is mainly due to insufficiently considering important spatio-temporal variables such as moisture and temperature gradients, and wood anatomical features during drying in the predictive modelling framework. To understand the influence of aforementioned variables, an experimental measurement set-up was developed consisting of heat-resistant load cells mounted in an oven to record the weight of the wood specimens dried at 100 °C. A camera was mounted outside the oven to acquire images of the end grain through the transparent door of the oven, as such enabling to monitor the occurrence of cracks. Furthermore, custom-made electrodes were inserted as well at different depths in the specimen to measure the electrical resistance, which relates to the moisture content. Specimens were also equipped with thermistors to record the temperature in the specimens at different depths from the surface. Heat-resistant wires from load cells, electrodes and thermistors were connected with data loggers outside the oven. Three specimens of basralocus (Dicorynia guianensis) were selected, measuring 50 mm wide, 50 mm thick and 120 mm long. The results show that it is possible to real-time monitor average moisture loss, local moisture content gradient represented by the electrical resistance gradient and wood temperature gradient as a function of time. These temporally resolved gradients can be related to the occurrence of cracks. This methodology enables to unveil interrelationships between the measured variables and changes of the wood anatomical structure, of utmost importance for the fine-tuning of kiln-drying schedules.